Over on his Hack & Slash blog, blogger C argues that realism and “making sense” are terrible and always make games worse.  Far from being terrible, realism and making sense are essential for fun. Even in such an abstract game as Tetris, the pieces have to “realistically” fit where blocks of that shape in the real world would fit, and realistically maintain their shape as you rotate them, and the controls have to “make sense” in that the left arrow moves them left, the right arrow moves them right, and the down arrow moves them down. None of those things had to be true, since it’s just a computer game, but the fact the code makes them true, makes them behave the way you’d expect, is part of what makes the game fun. Games are learning tasks, and learning things that don’t make sense is frustrating and un-fun.

Here’s part of  Hack & Slash: On why Realism and “Making Sense” are Terrible

Making things more realistic ruins games. Changing things to have them “make sense” destroys fun. I’ve written and designed computer games before and the most important lesson I learned from those experiences was to design fun mechanics and make the game about that fun. Jeff Vogel talks about it here. Every time someone suggested a way to make the game ‘more realistic’, it never failed detract from the game. Add armor damage and wear and tear on weapons causes tedium. Make the monsters fight each other causes endless messages and rooms full of dead creatures. How about at a table? Making people remember to eat, go to the bathroom and feed horses? You’ve insured that the players recite a list of items at various intervals. Sounds super fun, right? – C, the Hack & Slash blogger

C is confusing complexity with realism.  I agree with what C says later on completely, in that player agency is crucial, and that “In a game, an enjoyable activity comes from making choices with significant consequences.”  I even agree that the various activities in the game ought to be fun individually as they’re played.  But C draws the wrong conclusion from this. The problem with the things that C identifies as detracting from the games is that they represent a bad trade-off of extra complexity vs. the extra amount they help you in making choices with significant consequences, not that realism and making sense are bad things for a game.  The problem isn’t that they are realistic, the problem is that they are minor extra decisions (when to stop to eat, how much food to carry) that require constant bookkeeping to figure out the consequences.

If C’s diagnosis were right, and all that was necessary was that the activity in the game be fun in and of themselves, then you could improve any game just by substituting some more fun sub-game for any less fun activity.  Poker is more fun than rolling a die to see if you get a high number, so just resolve combat by playing poker.  Othello can substitute for social interaction.  Etc.  The problem with this approach is that it doesn’t make any sense, and it isn’t realistic.  It’s not realistic in the specific sense that the decisions that you make in gaining victory in the sub-system are nothing at all like the decisions you would make performing the activity that’s being abstracted; Tetris is a blast, but a computer racing game where how well your car was doing in the race was a factor of how well you were playing Tetris (turning the wheel left and right to move the blocks, shifting to rotate them, stepping on the pedal to drop the block) would be stupid.  You might even have fun playing the Tetris part,  but the car moving around the track is just distracting window dressing.  The game is improved by adding realism to it, by having the car turn left or right based on turning the wheel and accelerate by pressing the pedal.  It might be further improved by adding a shift lever that changes the way the virtual car responds… or that may be a step too far, and whether it’s an improvement or not can depend on the player’s tolerance for complexity and steep learning curves.  What’s not true is that this added realism would detract from the game, and every step towards more realism makes the game worse.  Note further that from a purely game rules point of view, it doesn’t matter at all whether turning the wheel right makes the car turn to the right, or to the left, or for that matter whether the facing of the car is controlled by the foot pedal, while the wheel controls acceleration.  As pure game elements representing decisions and skills to be mastered, they are equivalent.  The reason that one particular choice is fun and any other choice is just goofy is that there is one choice that’s realistic in that it allows you to directly map your understanding of the real world into how things will work in the game world; this is the essence of empowering the player to make meaningful decisions rather than just arbitrary game-optimal ones.

Take a step back for a moment.  Given that the point of the game is having the players make choices with significant consequences, what is at issue is how best to empower the players to make meaningful choices.  Often, if not always, the best way to do that is to make the game more realistic, to make the salient aspects make more sense. Empower the players to make good game choices by designing the game so that good choices based on real-world knowledge and reasoning come out to be good game choices.  On the flip side, for the love of Mike don’t go bananas and add meaningless choices or rules that have you doing twenty minutes of bookkeeping for every 10 seconds you get to think about making a choice: that’s bad design whether you add the rule in the name of realism, game balance, or any other desirable quality in a game.